Abstract: | Obtaining recommendations from guests is critical for Airbnb hosts to thrive in the peer-to-peer accommodation business. By extracting the dominant aspects of the Airbnb experience and the guests' sentiment ratings in online comments, this study examines the impact of the aspects and sentiment on guest actual recommendations. A novel mixed-method was adopted to analyse the online reviews of Airbnb guests in Los Angeles, USA. The text-mining results reveal that hosts, location, and amenities are the dominant aspects of the guests' Airbnb experience. Results show that the guests' sentiment ratings for hosts and amenities have significant impacts on the guests' recommendations, while location only influences the private-room guests' recommendations. This study contributes to the extant literature by offering an innovative methodological approach to building the link between experience aspects and guest actual recommendations. |